Build a trip from a brief

Here's an enquiry, exactly as it arrived:

"Couple, late February, 10 days, want the northern lights and a glacier walk, nice hotels but nothing flashy, not too much driving."

That's it. One sentence. No document, no supplier PDF, no template.

I'm Eric, from TripCreator, and this is the final episode of the series. Last time, the AI read a document someone else had written. This time there's no document at all, just a brief from a customer. Today is about building a trip from that brief, and about the one rule that makes it trustworthy.

Over the next twenty minutes I'll explain how it works, walk through that northern lights trip day by day, show you where every item in the draft comes from, share how to write a brief that gets a good draft, and then pull the whole series together.

Every travel planner knows the blank page. An enquiry comes in, and before you can send anything you have to build a first draft: pick the route, choose the hotels, place the activities, check the driving times and price it. For a 10-day trip, that's an hour or two before the customer has seen a thing.

And most enquiries don't book. So a lot of those hours go into drafts that never become trips.

The idea behind this feature is simple. Let the AI do the first draft, so your time goes into the trips that are worth refining.

Here's how it works. You paste the brief into TripCreator, the sentence from the email or your notes from the phone call, and you get a draft itinerary in your planner, laid out day by day.

For our couple, that means ten days in late February, arranged around the northern lights and the glacier walk, with comfortable hotels and short driving days. It's a real, editable trip in your planner, not a paragraph of suggestions.

But how the draft is built matters far more than how fast it arrives. So let's talk about where every item comes from.

There's a strict order of preference.

First, your own products. The AI looks in your product library before anywhere else: the hotels, tours and transfers you already sell, at your prices and with your margin. If you have a glacier walk in your library, the draft uses your glacier walk, from your supplier, at your price.

Second, real places. When the trip needs something you don't have in your library, like a particular waterfall, a viewpoint or a well-known restaurant, the AI can place a real place on the right day. But it doesn't invent a supplier or a price for it. It's a real location the guest can visit, and nothing more.

Third, placeholders. When nothing suitable is found, the draft gets a placeholder: a clearly marked gap, like "hotel, night four", for you to fill in. It doesn't make up a hotel. A placeholder is a normal, honest outcome, not a failure.

Let's look at what the draft for our couple might contain, and where each piece comes from.

Days one and two: arrival, and two nights at a comfortable hotel near the airport city. That hotel is one of yours, from your library, at your rate. The airport transfer is yours too.

Day three: a short drive along the coast. A well-known waterfall is placed on the route as a real place, with no price, because it's free to visit and you don't sell it. The hotel that night is yours.

Days four and five: two nights in the same place, because the brief said not too much driving. Your glacier walk is on day five, from your library, at your price.

Day six: a northern lights evening tour from your library, placed in a stretch of the trip where the hotels are away from city lights. The AI chose and placed it. It came from your library.

Day seven: a hotel placeholder. Nothing in your library matched that area for those dates, so the draft says so honestly, instead of inventing somewhere.

Days eight to ten: the drive back towards the airport, a final night in the city, and departure.

So the draft is mostly your own products, a couple of real places that make the days richer, and one clearly marked gap. In the next ten minutes, you fill the gap, swap one hotel you know this couple would like better, and the trip is ready to send.

This brings us to the rule I mentioned at the start, and it's the most important thing in this episode.

The AI chooses and arranges things. It never writes prices or suppliers of its own.

Technically, the AI is given a list of real things it's allowed to use, from your library and from real places, and it can only pick from that list. Then every item in the draft is rebuilt from our own records. The name, the supplier, the price and the category all come from your data, not from the AI's wording. So even if the AI phrased something oddly, what lands in your trip is your actual product.

Why go to all that trouble? Because the alternative is a draft that looks perfect and can't be delivered.

Picture an AI that writes a lovely itinerary featuring a charming boutique hotel at a very reasonable price. The customer loves it and says yes. Then you discover the hotel doesn't exist, or it's not one of your suppliers, or the price was made up. Now you have a disappointed customer and an awkward phone call.

With TripCreator, every item in the draft can be traced back to something real: your product, a real place, or an honest gap. The trip you send is one you can actually deliver at the price you quoted.

It's the same principle as last episode. AI import reads and doesn't invent. Build-from-a-brief chooses and doesn't invent. Throughout TripCreator, the AI does the legwork, and your data stays the source of truth.

The draft is only as good as the brief, so here's how to write a good one. Think of it as briefing a capable new colleague.

Give the basics. When, how long, and who's travelling. "Late February, ten days, a couple."

Say what matters most. The one or two things the trip must include. "Northern lights and a glacier walk."

Describe the style. The level of comfort and the feel. "Nice hotels but nothing flashy."

Describe the pace. How much moving around the guests want. "Not too much driving."

Mention what to avoid. If the guests have said no to something, say so.

You don't need to write an essay. Our example was one sentence, and it covered nearly all of that. But the more of the guest's real wishes you include, the closer the first draft lands.

And once the draft arrives, here's what to check before you send anything.

Check the shape of the trip. Does the route make sense, and is the pace what the guest asked for?

Fill the placeholders. Each one is a gap the AI couldn't fill from your library. Decide what goes there.

Look at the real places. Are they the stops you'd actually recommend? Swap any you wouldn't.

Then add your own touch. The sentence at the top of each day, the reason you've put the rest day where it is, the dinner you'd fight for. That's the part only you can write, and it's what wins the booking.

Once the draft is in your planner, it's a normal trip. Everything from the last nine episodes applies.

Swap a hotel, add a day, fill in the placeholders. Your markup is applied, so the price is right. Send it as a live itinerary page, the kind we talked about in episode four. Let the customer pick their optional extras. Take the deposit online, as in episode eight. Then operate the trip, with booking requests, vouchers and a guide in the guest's pocket.

The AI wrote the first draft. Everything after that is the same TripCreator you've been hearing about.

Think about what this changes for your team. The first hour on an enquiry used to be assembly. Now it's judgement: is this the right route for this couple? Would they prefer two nights in the north? Should the glacier walk move to day three, before the weather turns?

That's the work that actually wins bookings, and it's the work your best planners are good at. The AI takes the assembly off their plate.

It also changes what's worth doing. When a first draft takes minutes rather than hours, you can afford to answer every enquiry properly, including the ones that might not book. Some of those will.

Here are the questions operators usually ask about building from a brief.

"Will it replace my planners?" No. It replaces the assembly at the start, not the judgement. Your planners still shape the trip, write the story and talk to the customer.

"What if my product library is small?" Then you'll see more real places and more placeholders in the drafts, and that's honest. As your library grows, through your own work and through AI import from episode nine, the drafts use more of your own products.

"Does it use other companies' products?" No. It draws on your own library, plus real places with no supplier or price attached. Anything priced in the draft is yours.

"Can it get things wrong?" It can make choices you wouldn't, like a stop you don't rate or a pace you'd adjust. That's what your review is for. What it can't do is invent a hotel or a price, because it can only choose from real things and every item is rebuilt from your data.

This is the last episode, so let me pull the series together.

In episode one, we counted the tools behind one trip and found the spreadsheet tax: the cost of moving the same information between them by hand. The answer was to build each trip once, from your own products, and let that one version drive everything else.

In episode two, we saw that travel businesses are different on top and the same underneath, and that most of them are two or three kinds of business at once.

Then we went segment by segment. DMCs, answering the Tuesday brief by Wednesday morning, with every group size priced and every version kept. Bespoke sellers, making the proposal a page the guest wants to open. Series operators, making one change instead of twenty-four. Self-drive operators, turning the itinerary into the guide. And agencies and groups, keeping the margin visible and sharing what should be shared.

In episode eight, we let customers book at 11pm, into the same list of bookings as everything else. And in the last two episodes, we let the AI do the copying and the first draft, reading documents and choosing from real things, without ever inventing a price.

That's TripCreator in one breath. Build once. Present beautifully. Sell around the clock. Operate without spreadsheets. And let the AI do the legwork, while your data stays the source of truth.

To show how the brief shapes the draft, here are three quite different enquiries and what you'd expect from each.

The first is a family. "Two adults and two children, aged 8 and 11, one week in July, they love animals and swimming, and the parents want at least one evening to themselves." Here the draft leans towards fewer hotel changes, shorter travel days, and activities from your library that suit children. The evening to themselves is a nice test. If you have a babysitting service or a hotel with a kids' club in your library, the draft can use it. If you don't, you'll probably see a placeholder or a note, and that's your cue to add the personal touch.

The second is a small group. "Twelve colleagues, four days, a mix of team activities and good food, based in one city." Here the draft stays in one place, with the hotel from your library, and fills the days with group-friendly activities and restaurants you work with. Because it's a group, you'll want to check the group pricing and the room mix afterwards, exactly as we discussed in the DMC episode.

The third is an anniversary. "Our 25th anniversary, ten days, somewhere warm, we want one truly special night." Here the brief is about feeling more than logistics. The draft gives you a sensible structure with your best warm-weather hotels, and leaves the "truly special night" to you. It might use the finest hotel in your library for that night, but the real magic is what you add: the private dinner, the note in the room, the view you know about.

In all three cases, the pattern is the same. The AI handles structure and selection from what's real. You handle the moments that make the trip personal.

Let me end the feature part of this series with a broader thought about AI in travel businesses, because there's a lot of noise about it.

You'll hear claims that AI will plan entire trips on its own, that travel advisors will disappear, and that customers will simply ask a chatbot for a holiday. Some of that will happen at the edges. But for the kinds of businesses listening to this series, I think the reality is more useful and less dramatic.

AI is very good at the parts of travel work that are about moving and organising information. Reading a supplier's document. Laying out a first draft. Picking sensible options from a list. Those are exactly the parts that have always eaten your team's time, and they're the parts where TripCreator uses AI.

AI is not good at the parts that make your business worth choosing. Knowing which guide makes a group laugh. Knowing that a particular couple will love the quiet ryokan more than the famous one. Knowing which supplier will help you at midnight when something goes wrong. Standing behind the price you quoted.

That's why the rule of this episode matters so much. An AI that invents things would be competing with your judgement and your reputation. An AI that only chooses from real things, and leaves honest gaps, works for you. It takes the assembly off your plate and gives you back the time for the parts only you can do.

The businesses that do well over the next few years won't be the ones that hand everything to AI, or the ones that ignore it. They'll be the ones that use it for the legwork, and invest the time it gives back in the things their customers actually remember.

Today's takeaway: paste the brief, get a draft trip built from your own products, real places and honest gaps, and spend your time where your judgement matters. Write the brief the way you'd brief a colleague, check the draft the way you'd check their work, and add the touches only you can add. The AI chooses and arranges. It never invents.

If you'd like to see this with your own eyes, start a trial and paste in a brief of your own. The link is in the show notes. Or book a demonstration, bring a real enquiry from last week, and we'll build it live from the brief, so you can compare the result with what your team sent.

Thank you for listening to all ten episodes. If they were useful, pass them on to a colleague who's still paying the spreadsheet tax.

I'm Eric, from TripCreator.